SA1 Deliverable Review Form

Reviewed document with minor corrections.

General comments

The structure of the document is good and the tabular form of the plan would help the reader in understanding what will come. The plan however is quite dry and doesn't provide enough details for a real comprehension of the ongoing developments, see comment #10.

There are some issues with the document formatting, which should be addressed. Moreover the terminology should be reviewed and made uniform (e.g GLUE vs Glue, Compute Service vs Computing Service, etc.)

FG 21/11/10: on version 7 I am not completely happy with the v7 and the answers i got, nevertheless i can live with it. I leave it up to Laurence to decide if he considers any of my open issues worth to be taken into account.

If Laurence doesn't plan to make other changes, just let me know and I'll mark the still open issues as "not addressed" and we can close this review (from my point of view).

FG 29/11/10: on version 8 The new version doesn't address any of the open issues of my review and no further comments have been added in TWiki, so I've just marked them as "not addressed".

Additional recommendations (not affecting the document content, e.g. recommendation for future work)

Detailed comments on the content

Note 1: The reviewers must list here any observation they want to track explicitly and that require interaction with the authors
Alternatively all changes must be listed in the document itself using Word change tracking features (if you use Word)
Note 2: These comments have to be explicitly addressed by the authors and the action taken must be clearly described

N°	Page	Section	Observations	Is Addressed?
1	5	1.1	The Purpose is not the abstract. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
2	5	1.2	Include here the contents of the Executive Summary. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
3	5	1.4	Include the title of the references, not just the URL, where applicable. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
4	5	1.5	Add a reference to DNA1.1 or just copy the boilerplate included in the deliverable template. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010 You could add an item in the Table of References. -- FrancescoGiacomini - 20-Nov-2010
5	5	1.6	Remove the third column of the table of definitions, sort alphabetically the entries, format them in the same way (e.g. some entries are in bold), all the entries must have a definition (e.g. what is EGIIS?), there must be only one entry per raw. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010 You could add a vertical line on the right to close the table.
6	7	2	Expand the executive summary to include a summary of what is written in the rest of the document, e.g. mention some highlights from the development plan. Move the sentences on the structure of the document to the Document Organisation section (1.2). -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
7	11	3.3	"external source" -> "external destination"? -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
8	11	3.5	There is no mention of existing monitoring tools adopted by several services. From the top of my head I can mention the WMS (WMSMonitor) and FTS, but there are probably others. -- FrancescoGiacomini - 25-Oct-2010 Rejected. What is written reflects what is in the DoW. There may be custom service interfaces but we do not have a generic solution for EMI. -- LaurenceField - 03-Nov-2010 This deliverable is also about the state of the art, which for me means that all the existing known tools should be mentioned. If they are or could become a generic solution I guess will be the subject of the part about the workplan for the second year, since the DoW foresees a milestone on Service Monitoring and Management at M24 (MJRA1.15 - Service monitoring and management). -- FrancescoGiacomini - 20-Nov-2010 This deliverable is about status and plans. We have in the plan "Investigate Requirements", which includes providing the state-of-the-art. Although existing services may provide serivce monitoring and management features, this is not the focus of the topic. It is about integrating services into both Fabric and Infrastrucutre tools. -- LaurenceField - 08-Dec-2010
9	12	3.6	That virtualization started in the sixties is irrelevant, even distributed computing started tens of years ago. The current contents of this section better fits in the second part of the document on the working plan. Here I would expect a (brief) presentation of the current experience by several of the EMI partners on the subject. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010 Ok for the removal of the sentence on virtualization that started in the sixties, but there is nothing on the state of the art from the EMI partners. Don't you think it would be useful to include something? Moreover the paragraph is numbered and there are a few typos (e.g. "the can" -> "they can", "to off" -> "to offer"). -- FrancescoGiacomini - 20-Nov-2010 I fixed the typos. Again investigating the state-of-the-art is part of the goal for the first year. -- LaurenceField - 08-Dec-2010
10	13	4	In general the plan is quite dry. A reader would expect that the "Details" raw of each entry would contain at least what needs to be done and why, i.e. what is the justification of each development. Moreover it would be useful to note what is the impact on other components, especially in other Technical Areas. Please, review the entries so that at least the above information is available. What does the "due date" refer to? if it's the date of official release, most of them are before the first EMI release is available. Does it mean that these developments will be released in the existing ARC, gLite and UNICORE distributions? And why some dates are marked "n/a"? It would be useful that each entry (table) be numbered, so that it can be referenced. More specific comments follow. -- FrancescoGiacomini - 25-Oct-2010 Rejected. The other reviewesr did not make a similar comments and were happy with the content. -- LaurenceField - 03-Nov-2010 The fact that the other reviewers didn't raise any objection in this sense doesn't mean that you can ignore mine. All comments have to be addressed, not just their intersection set. You are free of course to reject my observation, but please use some argument. -- FrancescoGiacomini - 20-Nov-2010 From what I understand, the output of the Infrastructure Area and PTB is to define high-level objectives which can be turned in to concrete development tasks by JRA1. Thus there should be no "details" on the concrete development required. Many of the objectives are high-level investigations. The information that you suggest should be there is infact the output of these investigations -- LaurenceField - 08-Dec-2010
11	13	4.1.1	It's not clear what the "harmonization of information system clients" is. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010 Ok, but "to low" -> "to lower"? -- FrancescoGiacomini - 20-Nov-2010
12	15	4.2.1	For APEL there is no mention of work to move to ActiveMQ or to adopt the OGF-UR standard. Why? -- FrancescoGiacomini - 25-Oct-2010 Rejected. There is not a specific requirement for either of these. It is messaging and standards. -- LaurenceField - 03-Nov-2010 I don't understand the objection. Can you elaborate a little bit? do you mean that what concerns messaging and standards should not be included in this document? if so, why are those two items mentioned for DGAS? -- FrancescoGiacomini - 20-Nov-2010 The comment is there is no mention of work to move to ActiveMQ or to adopt the OGF-UR standard. We do not have a requirement to move to ActiveMQ or to adopt the OGF-UR, therefore it is not in this years plan. -- LaurenceField - 08-Dec-2010
13	16	4.2.2.1	Remove the title of this section. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
14	16	4.2.2.1	In the "Details" of "New Usage Record producer" point (1) is mentioned, but points are not numbered (although they should, see general comment #10). -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
15	16	4.2.2.1	Entries 3 and 4 have the same title. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010
16	16	4.2.2.1	In point 3 storage accounting is mentioned. This is a big issue and one of the major objectives of the project in this area. Is there an agreement/specification on how it has to be done? -- FrancescoGiacomini - 25-Oct-2010 Rejected. A task force has been set up to solve the issue. This was unknown at the time of writing. -- LaurenceField - 03-Nov-2010 Exactly my point. How can you plan the work if a task force has just been set up? -- FrancescoGiacomini - 20-Nov-2010 According to the technical development plan, defining the Storage Accounting record is a Data Area activity and the implementation of such (a cross area activity) is not scheduled until the second year. -- LaurenceField - 08-Dec-2010
17	17	4.3	It doesn't make much sense that the activity on messaging is limited to a survey and a collection of requirements when a lot of activity has already been done on the subject, including the deployment of a whole infrastructure based on ActiveMQ and its adoption by a number of services, e.g. all the accounting, which is mentioned just in the previous paragraph. -- FrancescoGiacomini - 25-Oct-2010 Addressed. There is a new objective on Messaging to provider guidelines earlier. -- LaurenceField - 03-Nov-2010
18	18	4.5	In the point about virtualization, in "ensure that EMI components can support virtualization" what does "support" mean? Please elaborate, e.g. on the GLUE schema, on the job description, on the involved components, etc. -- FrancescoGiacomini - 25-Oct-2010 Addressed. Please verify. -- LaurenceField - 03-Nov-2010 Is the change limited to "virtualization" -> "virtualized deployment"? is this workplan item just about ensuring that a certain service can be deployed within a virtual machine rather than on a native platform? if so, aren't we missing something, i.e. the possibility to run jobs (or whatever you want to call them) within a virtualized environment, which is a more interesting problem to solve? -- FrancescoGiacomini - 20-Nov-2010 The possibilty to run jobs in a VM is a Compute Area activity and scheduled for the second year. -- LaurenceField - 08-Dec-2010

Any other modification, spelling or grammatical corrections, etc. must be done directly in the document using tracked changes or similar mechanisms that allows the authors to identify which correction is suggested.

-- FloridaEstrella - 14-Oct-2010